“The future of asset management is to combine fundamental approaches with quantitative techniques,” explains Denis Panel, who is Chief Investment Officer and Head of MAQS at BNPP AM, as well as being Professor of Asset Management on the MBA programme at the top-ranked HEC business school in Paris. He continues, “We have introduced new quantitative signals, based on macro or monetary policy factors, alongside discretionary fundamental input which comes from the investment committee, for timing beta in equities, and will soon do so for fixed income. Neither part is an overlay and sometimes the two views are opposing. For instance, our discretionary managers have strong conviction on US equities, but our quantitative models are negative on them. The objective is for the quantitative models to help mitigate the behavioural biases that can come from discretionary management.” 

Discretion also informs market selection in BNPP AM’s purely quantitative strategies. Japanese government bonds were removed from one of its models after the Bank of Japan announced measures to strictly control long term yields in 2017.

“The latest innovation is a new optimiser which has been developed for multi-factor allocation, based on an in-house research paper,” he adds. “It is more stable than a traditional Markowitz mean/variance optimiser, resulting in more stable, robust and intuitive results. It has been used for strategic asset allocation for some time and is now being used for tactical asset allocation as well, on multi-asset and benchmark products.” By contrast, the ISOVOL range rebalances exposures based on judgement and recent market volatility, to maintain a constant volatility target, using GARCH techniques to forecast volatility. Unlike risk parity, ISOVOL does not imply equal risk contributions per asset class and both long biased and long/short MAQS products have had lower weightings in bonds than many risk parity strategies.